Doxorubicin Improves Cancer Cell Targeting by Filamentous Phage Gene Delivery Vectors

Merging targeted systemic gene delivery and systemic chemotherapy against cancer, chemovirotherapy, has the potential to improve chemotherapy and gene therapy treatments and overcome cancer resistance. We introduced a bacteriophage (phage) vector, named human adeno-associated virus (AAV)/phage or AAVP, for the systemic targeting of therapeutic genes to cancer. The vector was designed as a hybrid between a recombinant adeno-associated virus genome (rAAV) and a filamentous phage capsid. To achieve tumor targeting, we displayed on the phage capsid the double-cyclic CDCRGDCFC (RGD4C) ligand that binds the alpha-V/beta-3 (αvβ3) integrin receptor. Here, we investigated a combination of doxorubicin chemotherapeutic drug and targeted gene delivery by the RGD4C/AAVP vector. Firstly, we showed that doxorubicin boosts transgene expression from the RGD4C/AAVP in two-dimensional (2D) cell cultures and three-dimensional (3D) tumor spheres established from human and murine cancer cells, while preserving selective gene delivery by RGD4C/AAVP. Next, we confirmed that doxorubicin does not increase vector attachment to cancer cells nor vector cell entry. In contrast, doxorubicin may alter the intracellular trafficking of the vector by facilitating nuclear accumulation of the RGD4C/AAVP genome through destabilization of the nuclear membrane. Finally, a combination of doxorubicin and RGD4C/AAVP-targeted suicide gene therapy exerts a synergistic effect to destroy human and murine tumor cells in 2D and 3D tumor sphere settings.


Introduction
Gene therapy has been attempted against cancer for the past 28 years, with more than 67% of clinical trials of gene therapy designed to treat cancer patients [1]. However, progress has been hindered, mostly by a lack of tumor-selective gene delivery vectors efficient via clinical systemic routes and by problems with repeated vector administrations [2,3]. Currently, the most common vectors forcing dose-limiting treatments and combinations with other anticancer agents [29]. We found that doxorubicin improved gene delivery by RGD4C/AAVP without affecting the vector selectivity for cancer cells, subsequently enhancing tumor cell killing. We also investigated the effects on vector cellular trafficking, since doxorubicin can alter the intracellular trafficking of eukaryotic viral vectors [30], and uncovered that destabilization of the nuclear envelope is one mechanism by which doxorubicin might enhance gene delivery.

Doxorubicin Increases Targeted Gene Transfer by RGD4C/AAVP in Rat Glioma (9L) and Human Melanoma (M21) Cancer Cells In Vitro
First, we determined the concentrations of doxorubicin that give the optimal gene delivery efficiency by RGD4C/AAVP in 9L and M21 tumor cells. Cells were transduced with vector carrying the firefly luciferase, Luc, reporter gene (RGD4C/AAVP-Luc) in combination with increasing concentrations of doxorubicin ranging from 0.5 to 8 µM for 24 h. At day three post-vector transduction, we found that 0.5 µM for 9L and 0.6 µM for M21 cells were the optimal doxorubicin concentrations that generated the highest Luc expression by the vector ( Figure 1A). Next, we used these optimal doxorubicin doses and carried out a broader investigation of the effect of doxorubicin on gene delivery by RGD4C/AAVP by using two different reporter genes performing time course gene delivery experiments and testing the efficacy both in 9L and M21 tumor cells. First, we used vectors expressing a reporter gene of the green fluorescent protein (RGD4C/AAVP-GFP). At day four post-vector transduction, we observed more GFP expression in 9L and M21 tumor cells treated with a combination of RGD4C/AAVP-GFP and doxorubicin than in cells transduced with the RGD4C/AAVP-GFP vector alone ( Figure 1B). To further evaluate doxorubicin effects on gene delivery levels, we repeated the GFP reporter gene expression experiments in the presence of increasing doxorubicin concentrations on the UW228, as an in vitro model of human medulloblastoma, which is the most common brain cancer in children. Herein, we used a fluorescence-activated cell sorting (FACS) analysis of GFP expression in UW228 cells and showed a substantial increase of GFP-expressing cells by doxorubicin, reaching 55% GFP-positive UW228 cells in the presence of 8-µM doxorubicin as compared to 11% of GFP-expressing cells treated with the vector alone ( Figure 1C). Next, we used the RGD4C/AAVP-Luc vector and monitored luciferase expression over a time course ( Figure 1D). Our data showed an increased luciferase expression by RGD4C/AAVP-Luc in 9L and M21 cells over time in the presence of doxorubicin as compared to treatment with the vector alone ( Figure 1D). For example, at day four post-treatment, an analysis of Luc transgene expression showed that the combination treatment resulted in a~5.3 and~12-fold increase in Luc expression in 9L and M21 cells, respectively, compared to the RGD4C/AAVP-Luc vector alone. Moreover, in 9L cells, initiation of Luc expression occurred as early as day two post-vector transduction in the presence of doxorubicin ( Figure 1D). Importantly, no Luc expression was detected in cells treated with the control nontargeted AAVP-Luc vector (lacking in RGD4C, fd-Luc) alone or in combination with doxorubicin, which shows that doxorubicin does not affect targeting of the RGD4C/AAVP vector and its selectivity for integrin-expressing cells. of optimal Dox doses in 9L and M21 cells using a luciferase assay. The 9L and M21 cells were transduced with the targeted RGD4C/AAVP-Luc (luciferase) vector carrying a luciferase (Luc) reporter gene, in combination with increasing concentrations of Dox ranging from 0.5 to 8 μM for 24 h. To evaluate gene delivery, a luciferase assay was performed at day 3 post-transduction. The results were normalized to the amount of protein, as determined by the Bradford assay, and shown as relative luminescence units (RLU)/μg of protein. (B) Evaluation of the gene expression in 9L and M21 cells in the presence of optimal Dox concentrations (0.5 μΜ for 9L and 0.6 μΜ for M21 cells). Tumor cells Figure 1. Doxorubicin (Dox) boosts targeted gene delivery by the RGD4C/ adeno-associated virus phage (AAVP) vector in rat glioma (9L) and human melanoma (M21) cancer cells. (A) Determination of optimal Dox doses in 9L and M21 cells using a luciferase assay. The 9L and M21 cells were transduced with the targeted RGD4C/AAVP-Luc (luciferase) vector carrying a luciferase (Luc) reporter gene, in combination with increasing concentrations of Dox ranging from 0.5 to 8 µM for 24 h. To evaluate gene delivery, a luciferase assay was performed at day 3 post-transduction. The results were normalized to the amount of protein, as determined by the Bradford assay, and shown as relative luminescence units (RLU)/µg of protein. (B) Evaluation of the gene expression in 9L and M21 cells in the presence of optimal Dox concentrations (0.5 µM for 9L and 0.6 µM for M21 cells). Tumor cells were plated in 48-well plates and transduced with the targeted RGD4C/AAVP-green fluorescent protein (GFP) (RGD4C-GFP) vector carrying a GFP reporter gene in the presence or absence of Dox. GFP expression was evaluated by fluorescent microscopy at day 4 post-vector transduction. GFP expression is also displayed as the average of GFP-positive cells in five fields of view of treated 9L and M21 cells. Original magnification ×200 for GFP and ×40 for brightfield images. (C) Fluorescence-activated cell sorting (FACS) analysis of GFP expression in UW228 medulloblastoma cells at day 5 post-transduction with RGD4C-GFP or nontargeted (fd-GFP) vectors with varying concentrations of Dox ranging from 0.08 to 8 µM for 24 h. (D) Evaluation of the effect of Dox on gene delivery over a time course. The 9L and M21 cells were plated in 48-well plates and transduced with the targeted RGD4C/AAVP-Luc (RGD4C-Luc) vector or control nontargeted AAVP-Luc (fd-Luc) in serum-free medium in the presence or absence of Dox 0.5 µM and 0.6 µM for 9L and M21 cells, respectively. Luciferase measurement assays were performed at various time points from days 1 to 4 post-vector transduction and normalized to protein concentrations. Results are shown as RLU per 1 µg of protein. All experiments were repeated twice, in triplicates, and the results shown are representative of one experiment. ** p < 0.01 and *** p < 0.001.

Doxorubicin Drug Treatment Boosts Cancer Cell Death by AAVP-Mediated Suicide Gene Killing
We sought to investigate whether the enhanced vector-mediated gene delivery by doxorubicin is translated into an improved targeted killing of cancer cells by the vector carrying a therapeutic gene. We used the RGD4C/AAVP-HSVtk vector transferring the gene for HSVtk mutant SR39 [31], which kills cells in the presence of GCV. The HSVtk enzyme phosphorylates prodrug nucleoside analogs such as GCV and converts them into nucleoside analog triphosphates, which are then incorporated into the cellular genome, inhibit DNA polymerase and, subsequently, induce cell death by apoptosis [32]. Thus, 9L and M21 cells were treated with RGD4C/AAVP-HSVtk or control nontargeted vector fd-HSVtk in the presence or absence of doxorubicin. GCV (20 µM) was added after three days, and tumor cell death was evaluated at 0, 24, 48, 72 and 96 h following treatment with GCV. Data were normalized to the nontargeted fd-HSVtk vector that did not show any significant tumor cell death (Supplementary Figure S1). In both cancer cell lines, the combination treatment with doxorubicin and RGD4C/AAVP-HSVtk therapy resulted in greater cell killing compared to cells treated with RGD4C/AAVP-HSVtk or doxorubicin alone ( Figure 2). For example, 92 h after GCV addition, we measured 99.7% and 96.5% cell death of 9L and M21 by the combination treatment, respectively, compared to 87% and 58.3% cell killing by the vector alone in 9L and M21, respectively. The treatment with doxorubicin alone induced 82.4% and 90.9% death in 9L and M21 cells, respectively ( Figure 2). These findings prove that doxorubicin treatment is an effective strategy to improve RGD4C/AAVP-targeted gene therapy.

Evaluation of the Effect of Doxorubicin Treatment on Vector Cellular Trafficking
We set out to gain understanding into the mechanism of the enhanced RGD4C/AAVP-mediated gene delivery and tumor cell killing by doxorubicin. Therefore, we investigated whether doxorubicin alters vector trafficking through the various stages of gene transfer [33,34]. First, we monitored vector attachment to the cell surface in the presence of doxorubicin by quantifying the free phage present in the medium of cells by infection of host bacteria and colony counting as previously reported [21]. Our data showed that~37% of the phage input was recovered from the cell medium upon treatment with the RGD4C/AAVP-Luc vector, meaning that 63% of the added vector was attached to the tumor cell surface ( Figure 3A), though the combination with doxorubicin had no effect on the attachment of the RGD4C/AAVP-Luc vector to the surface of cancer cells ( Figure 3A). Moreover, we quantified 100% recovery of the nontargeted fd-Luc vector, proving no attachment to tumor cell surface. Next, internalization assays revealed that the combination of vector with doxorubicin did not increase entry of RGD4C/AAVP-Luc into cancer cells ( Figure 3B). These data prove that doxorubicin does not affect cell attachment and entry of the RGD4C/AAVP particles into cancer cells.
We also investigated the vector genome localization in the nucleus, since doxorubicin has been reported to delay proper chromosome condensation and nuclear envelope formation during mitosis [35], which could result in increased transport through the nuclear pores. Thus, we evaluated the nuclear accumulation of the AAVP genome as we previously reported [36] in the presence or absence of doxorubicin by quantifying the ITRs. We used these ITR cis genetic elements as targets for quantification, because they have been used to quantify the AAV genome copies (GC) in cells and because gene expression by AAVP is driven by an ITR-flanked transgene cassette or recombinant AAV genome [5,6]. Hence, nuclei were extracted from cells to perform PCR with primers reading within the AAV2 ITR domain and by using the original reported protocol [37] to semi-quantify the amount of ITRs in the nucleus. The data revealed that the combination of doxorubicin with the targeted vector resulted in a significant increase, 1.6-and 2.5-fold, of vector DNA in the nuclei of 9L and UW228 tumor cells, respectively ( Figure 3C). To further validate these data, we investigated the effect of doxorubicin on expression of the lamin B1 protein, a major component of the nuclear membrane, since doxorubicin was reported to induce cell cycle arrest of the G2/M phase during which the nuclear membrane is partially produced [38,39]. Immunofluorescence staining of cells following treatment with doxorubicin showed a lower staining of lamin B1 both in 9L and M21 cells compared to untreated cells ( Figure 4A). Next, Western blot analysis confirmed the drop of lamin B1 expression in 9L ( Figure 4B) and M21 cells ( Figure 4C) following treatment with doxorubicin in a dose-dependent manner and with the most decrease in lamin B1 obtained with the optimal doses of doxorubicin ( Figure 4B,C).

Evaluation of the Effect of Doxorubicin Treatment on Vector Cellular Trafficking
We set out to gain understanding into the mechanism of the enhanced RGD4C/AAVP-mediated gene delivery and tumor cell killing by doxorubicin. Therefore, we investigated whether doxorubicin alters vector trafficking through the various stages of gene transfer [33,34]. First, we monitored vector attachment to the cell surface in the presence of doxorubicin by quantifying the free phage present in the medium of cells by infection of host bacteria and colony counting as previously reported [21]. Our data showed that ~37% of the phage input was recovered from the cell medium upon treatment with the RGD4C/AAVP-Luc vector, meaning that 63% of the added vector was attached to the tumor cell surface ( Figure 3A), though the combination with doxorubicin had no effect on the attachment of the RGD4C/AAVP-Luc vector to the surface of cancer cells ( Figure 3A). Moreover, we quantified 100% recovery of the nontargeted fd-Luc vector, proving no attachment to tumor cell surface. Next, internalization assays revealed that the combination of vector with doxorubicin did not increase entry of RGD4C/AAVP-Luc into cancer cells ( Figure 3B). These data prove that doxorubicin does not affect cell attachment and entry of the RGD4C/AAVP particles into cancer cells.

Evaluation of Efficacy of Doxorubicin and RGD4C/AAVP Combination in a Three-Dimensional (3D) Multicellular Tumor Spheroid Model
We sought to investigate the combination of doxorubicin and the targeted vector in 3D tumor spheres commonly used as valid models to mimic the features of tumors in vivo [21,22,36]. Therefore, we first examined gene transfer using the vector carrying the GFP reporter gene and monitored GFP expression over time by microscopic observation of 9L and M21 tumor spheroids ( Figure 5). Only minimal GFP expression was detected in the 3D spheroids upon treatment with the targeted RGD4C/AAVP-GFP; however, the combination with doxorubicin significantly enhanced GFP expression from RGD4C/AAVP-GFP both in 9L and M21 tumor spheroids ( Figure 5). The treatment with the nontargeted fd-GFP vector did not result in any GFP expression.

Evaluation of Efficacy of Doxorubicin and RGD4C/AAVP Combination in a Three-Dimensional (3D) Multicellular Tumor Spheroid Model
We sought to investigate the combination of doxorubicin and the targeted vector in 3D tumor spheres commonly used as valid models to mimic the features of tumors in vivo [21,22,36]. Therefore, we first examined gene transfer using the vector carrying the GFP reporter gene and monitored GFP expression over time by microscopic observation of 9L and M21 tumor spheroids ( Figure 5). Only minimal GFP expression was detected in the 3D spheroids upon treatment with the targeted RGD4C/AAVP-GFP; however, the combination with doxorubicin significantly enhanced GFP expression from RGD4C/AAVP-GFP both in 9L and M21 tumor spheroids ( Figure 5). The treatment with the nontargeted fd-GFP vector did not result in any GFP expression.  Next, the application of HSVtk/GCV suicide gene therapy on rat 9L and human M21 tumor spheroids resulted in pronounced regression of the 9L and M21 ( Figure 6A) sphere volumes by the combination of doxorubicin with the targeted RGD4C/AAVP-HSVtk upon GCV treatment compared to individual treatments with the vector or doxorubicin alone. Subsequently, in 9L spheres, the measurement of cell viability showed that the combination of doxorubicin plus RGD4C/AAVP-HSVtk achieved higher tumor cell killing (~97%) than the RGD4C/AAVP-HSVtk vector or doxorubicin alone, which induced~92% and~65% cancer cell killing, respectively ( Figure 6B). In M21 spheres, the measurement of cell viability showed that the combination treatment achieved higher tumor cell killing (~91%) than the RGD4C/AAVP-HSVtk or doxorubicin alone, which induced~47% and~82% cancer cell killing, respectively ( Figure 6B). These findings establish that doxorubicin greatly enhances RGD4C/AAVP-mediated cancer gene therapy.
Next, the application of HSVtk/GCV suicide gene therapy on rat 9L and human M21 tumor spheroids resulted in pronounced regression of the 9L and M21 ( Figure 6A) sphere volumes by the combination of doxorubicin with the targeted RGD4C/AAVP-HSVtk upon GCV treatment compared to individual treatments with the vector or doxorubicin alone. Subsequently, in 9L spheres, the measurement of cell viability showed that the combination of doxorubicin plus RGD4C/AAVP-HSVtk achieved higher tumor cell killing (~97%) than the RGD4C/AAVP-HSVtk vector or doxorubicin alone, which induced ~92% and ~65% cancer cell killing, respectively ( Figure 6B). In M21 spheres, the measurement of cell viability showed that the combination treatment achieved higher tumor cell killing (~91%) than the RGD4C/AAVP-HSVtk or doxorubicin alone, which induced ~47% and ~82% cancer cell killing, respectively ( Figure 6B). These findings establish that doxorubicin greatly enhances RGD4C/AAVP-mediated cancer gene therapy. transduction with RGD4C-HSVtk or fd-HSVtk, the nontargeted vector, in the presence or absence of doxorubicin (0.5 μΜ for 9L and 0.6 μΜ for M21 spheroids). GCV was added to the spheres at day 5 post-vector transduction and renewed every 2 days. Images were taken at days 5 or 7 post-GCV treatment for 9L and M21 spheres, respectively. Original magnification ×20. (B) Evaluation of cell viability in 9L and M21 spheres at days 5 or 7 post-GCV treatment, respectively. Cell viability was

Discussion
Herein, we reported that the doxorubicin treatment of cancer cells resulted in increased RGD4C/AAVP-guided gene delivery, subsequently enhancing tumor cell killing by RGD4C/AAVP-HSVtk plus GCV both in 2D cell cultures and 3D tumor spheres. The combination of vector plus doxorubicin was more effective than the RGD4C/AAVP vector alone in reporter gene expression analyses than in tumor cell death experiments. However, the HSVtk/GCV strategy can produce a bystander effect, resulting in enhanced tumor cell death by the RGD4C/AAVP-HSVtk plus GCV [12]. Indeed, cells expressing the HSVtk can induce the death of surrounding cells via the gap junctional intercellular communications (GJIC), as previously reported [12,40]. Overtime, this "bystander effect" increases cell death by the RGD4C/AAVP-HSVtk, which may result in increased tumor cell death by the RGD4C/AAVP-HSVtk and GCV. Additionally, we used the mutant SR39 version of HSVtk that is more effective than the wild-type HSVtk [31]. Thus, the data of tumor cell death by HSVtk plus GCV should be partially proportional to the levels of HSVtk expression. This bystander effect could also explain the slight but insignificant effect of fd-HSVtk plus GCV (Supplementary Figure S1), which could be caused by unspecific phage uptake when using high vector amounts. Thus, we normalized our data to fd-HSVtk to subtract the effect of nonspecific uptake. Unspecific uptake of the nontargeted vector could also contribute to the slight increase of GFP-positive cells in FACS analysis experiments when treating UW228 cells with high amounts of fd-GFP in combination with a high doxorubicin concentration.
Nevertheless, these data are consistent with previous studies reporting that doxorubicin augments the transduction efficiency of airway cell lines by the recombinant rAAV vector [41]. Doxorubicin was also reported to increase rAAV-2 transduction in rat neuronal cell lines [30]; in that study, the authors suggested that doxorubicin improved the transduction efficiency by enhancing the nuclear translocation of rAAV. Indeed, topoisomerase II inhibitors, such as doxorubicin, have been reported to delay proper chromosome condensation and nuclear envelope formation throughout mitosis. This should increase the permeability of the nuclear membranes, subsequently facilitating nuclear transport [35]. Our data are consistent with these studies, as we found a greater vector genome in the nuclear fraction of cancer cells treated both with doxorubicin and the targeted RGD4C/AAVP vector compared to cells treated with the RGD4C/AAVP vector alone. Additionally, doxorubicin-mediated gene delivery enhancement was related to the levels of nuclear vector genome increase by this chemotherapeutic agent. Additionally, doxorubicin treatment resulted in a clear decrease of expression of the nuclear lamin B1 proteins involved in nuclear stability. During mitosis, the lamin B1 matrix is reversibly disassembled as the lamin proteins are phosphorylated [42][43][44]. Taking into account that the combination of RGD4C/AAVP with doxorubicin does not affect the vector attachment to the surface of cancer cells nor its internalization, our data postulate that doxorubicin may facilitate the nuclear translocation of the RGD4C/AAVP genome, subsequently enhancing transgene expression and consequent targeted cancer cell killing. Interestingly, we previously reported that most of drug combination treatments enhanced targeted RGD4C/AAVP gene delivery by altering the vector trafficking through obstacles to gene transfer vectors. Thus, doxorubicin action might also be mediated by its proteasome inhibitory activity [41]. Indeed, we previously reported proteasome degradation as an intracellular barrier to gene delivery by RGD4C/AAVP [24] and that proteasome-inhibiting drugs could be used to assist the phage to overcome this barrier and, subsequently, enhance gene delivery. We reported that proteasome-inhibiting drugs protect RGD4C/AAVP against proteasome degradation [24,36].
Moreover, genotoxic stress can increase AAV-mediated transduction efficiency by facilitating the conversion of single-to double-stranded DNA [45,46]. Therefore, another possible mechanism of doxorubicin-increased transduction efficiency of the RGD4C/AAVP vector could be that low-dose doxorubicin causes moderate DNA damage, resulting in activation of DNA repair enzymes, such as PARP-1(poly(ADP-ribose) polymerase-1) [29], which can replicate vector genomes. DNA repair enzymes could facilitate the conversion of the single-stranded genome of RGD4C/AAVP to double-stranded DNA, thus increasing transgene expression from the vector. This is consistent with our comet assay findings showing that the low doxorubicin doses that increased the transduction efficiency of our vector cause moderate DNA damage (Supplementary Figure S2). The comet assay is a single cell-based technique that allows the detection and quantification of DNA damage.
Together, these findings support our recent published studies reporting chemotherapy as an adjuvant to activate RGD4C/AAVP-HSVtk and GCV-targeted cancer gene therapy. Additionally, our current investigation has the potential to alter the clinical use of doxorubicin, and chemotherapy in general, and should influence systemic targeted cancer gene therapy by RGD4C/AAVP. Combinatorial treatment regimens represent a promising approach to overcome chemoresistance, which is a hallmark of cancer and should allow the reduction of chemotherapeutic drug administration to less toxic and cheaper doses. Moreover, both RGD4C/AAVP and doxorubicin are administered via clinical systemic routes (intravenous), suggesting that this combination treatment can be used to target distant tumors, a route that is applicable for metastatic disease. The treatment of metastatic cancer is heavily reliant upon systemically administered therapeutics. Metastases present a formidable challenge to successful cancer therapy and are responsible for as much as 90% of cancer-related deaths [47,48]. In this study, the proved efficacy of the combination between doxorubicin and RGD4C/AAVP-HSVtk plus GCV suicide gene therapy against M21 human melanoma provides a potential application of this chemovirotherapy against metastatic melanoma, known for its poor prognosis and less than 12 months median survival [49].
Doxorubicin is used to treat many different types of cancer and can be administered intravenously, like RGD4C/AAVP, over a course of several cycles of treatment over a few months. The dose of doxorubicin can vary according to indication and whether it is used as single therapeutic agent or in combination with other anticancer drugs. The toxicity of doxorubicin may be potentiated by other anticancer therapies and vice versa; thus, a combination might require dose adjustment [50,51]. A dose of 60-75-mg/m 2 body surface area is recommended every three weeks when doxorubicin is used as a single agent [50,51]. When doxorubicin is administered in combination with other antitumor agents with overlapping toxicity, the dosage of doxorubicin can be reduced to 30-60 mg/m 2 every three to four weeks [50,51]. However, bacteriophages are known for their long safety profile, since they have been administered to humans over a long time to treat bacterial infections [52] and were then approved in 2006 by the Food and Drug Administration for their use as antibacterial food additives [53]. Moreover, repeated intravenous administrations of RGD4C/AAVP carrying tumor necrosis factor alpha (TNFα), once a week over eight weeks, proved safe in dogs with natural cancers [20]. Additionally, the M13 phage displaying a peptide library, from which derived RGD4C/AAVP, was serially administered to cancer patients over a few weeks without any unwanted side effects [54]. Therefore, RGD4C/AAVP could be given safely and weekly in a combination without potential doxorubicin toxicity and might not require adjustment of the doxorubicin dosage and the interval between doxorubicin administrations.

Production, Purification and Titration of AAVP Vectors
Phage vectors were designed as previously reported [6] by inserting AAV2 transgene expression cassettes into the fUSE5 plasmid and expressing reporter or therapeutic genes under the control of the CMV promoter. Then, production and purification of AAVP phage particles were carried out as previously described [6] from the supernatant of host bacteria, followed by filtration through 0.45-µm filters. The titration of AAVP vectors was performed by colony counting following infection of the host bacteria, then expressed as bacterial transducing units (TU/µL), as previously reported [6].

In Vitro Cell Transduction by AAVP Vectors
The 9L and M21 cells growing in 48-well plates were incubated with targeted RGD4C/AAVP or control nontargeted AAVP vectors (1 × 10 6 TU/cell when using HSVtk vectors, 2.5 × 10 5 TU/cell for GFP vectors and 1 × 10 4 TU/cell when testing Luc vectors) in serum-free medium. Doxorubicin was added to 9L and M21 cells at the same time with AAVP vectors. In FACS analysis experiments for GFP expression, the UW228 cells were incubated with 1 × 10 6 TU/cell of RGD4C-GFP or nontargeted fd-GFP vectors using the same transduction protocol used for 9L and M21 cells. Following 4 h of transduction, 350 µL of complete medium containing doxorubicin were added to make up a total volume of 500 µL per well. Twenty-four hours after doxorubicin treatment, the medium was replaced with doxorubicin-free complete medium.

Reporter Gene Assays
Steady-Glo luciferase assay (Promega, Southampton, UK) was used to quantify luciferase expression that was then normalized to cell lysate proteins determined by the Bradford assay (Sigma, Dorset, UK). Luciferase expression was presented as relative luminescence units (RLU) per 1 µg of protein. A Nikon Eclipse TE2000-U fluorescence microscope (Tokyo, Japan) was used to monitor GFP expression.

Determination of Tumor Cell Killing In Vitro
Cancer cells were treated with 20-µM GCV (Sigma, Dorset, UK) at day 3 post-transduction with vectors expressing the HSVtk in the presence or absence of doxorubicin. Cell death was evaluated at 0, 24, 48, 72 and 96 h post-treatment with GCV by using the trypan blue exclusion methodology.

D Model of Multicellular Tumor Spheroid Culture and Treatment
The 9L and M21 multicellular tumor spheroids were established following 48 h culture of the tumor cells in 96-well ultra-low attachment surface plates (Corning, Nottingham, UK). Vector transduction was performed for 24 h. In combination treatment experiments, similar doxorubicin concentrations used for 2D cultures were applied for the tumor spheroids (0.5 µM for 9L and 0.6 µM for M21 spheroids). In the spheroid transduction experiments using vectors carrying the HSVtk gene, GCV (20 µM) was added to the spheroids on day 5 post-vector treatment and renewed every 2 days. The spheroid cell viability was measured by using the CellTiter-Glo assay (Promega, Southampton, UK).

Attachment Assay
The 9L cells were seeded in 48-well plates and grown for 48 h to reach 70-80% confluence. Cells were incubated with targeted and control nontargeted vectors carrying the Luc gene (1 × 10 6 TU/cell) in serum-free medium in the presence or absence of doxorubicin (0.5 µM). After 1-h incubation on ice, the supernatants were harvested to quantify AAVP, as previously described [6].

Internalization Assay
Internalization assay on 9 L tumor cells was done as we previously reported [21]. Briefly, targeted and control nontargeted vectors carrying the Luc reporter gene were incubated with 9L cells at 10 4 TU/cell at 37 • C in serum-free medium in the presence or absence of doxorubicin over a time course of 1, 2 and 4 h. To stop endocytosis, cells were cooled on ice, washed with PBS, then trypsinized to remove surface-attached phage and pelleted by centrifugation. Intracellular phage vectors were stained with a polyclonal rabbit anti-phage antibody (1:1000, Sigma, Dorset, UK) and a secondary goat anti-rabbit AlexaFluor 647 (dilution 1:500, Life Technologies, Renfrew, UK) antibody. FACS analysis was performed to measure the mean fluorescence intensity, and the results were analyzed by FlowJo software version 10.5 (TreeStar, Ashland, OR USA).

Nuclei Extraction
The 9L cells were first treated with targeted RGD4C/AAVP or control nontargeted vectors carrying the Luc reporter gene in the presence or absence of doxorubicin. At day 4 post-vector transduction, cells were harvested, and the nuclei were extracted, as previously reported [36].

Semi-Quantitative PCR Analysis
PCR semi-quantification of the ITR vector domain in the nucleus, in the presence or absence of doxorubicin, was performed as previously reported [37]. ImageJ software version 1.51j8 (NIH, Bethesda, MD USA) was used to quantify the PCR band intensity; then, data were normalized to GAPDH.

Comet Assay
Slides were immersed in 1% agarose in dH 2 0 at 55 • C and the agarose allowed to solidify. Cell suspension was pipetted and spread onto the first agarose layer using a coverslip. Second and third layers of melting agarose were added. Following solidification of the top layer of agarose, the slides were immersed in cold lysis buffer (2.5M-NaCl, 0.1-M Na 2 EDTA, 0.01-M Tris-HCl and 1% N-lauroylsarcosine sodium salt). The pH was adjusted to 10, followed by addition of 1% Triton X-100 and 10% DMSO before use. The slides were stored at 4 • C for at least 1 h to lyse the cells and to allow the unfolding of DNA. The slides were removed from the lysis buffer, drained and placed in a gel electrophoresis machine side-by-side. The tank was filled with fresh cold electrophoresis solution (1-mM Na 2 EDTA and 0.3-M NaOH) at 4 • C and pH 12.8. Electrophoresis was performed at 300 mA for 20 min at 4 • C in the dark to allow damaged DNA to migrate away from the nucleus, resulting in the formation of a comet "tail" whose size is proportional to the damage of DNA. Finally, at the end of electrophoresis, slides were stained with 6-diamidino-2-phenylindole (DAPI, 5 µg/mL) in mounting medium in order to view individual nuclei using a fluorescent microscope. Images of randomly selected nuclei were captured and analyzed using ImageJ software.

Immunofluorescence Staining for Lamin B1
Cells were grown on coverslips, then treated with doxorubicin for 24 h. Next, cells were washed with PBS and fixed in 4% paraformaldehyde for 15 min at room temperature. After fixation, cells were treated for 5 min with 50-mM ammonium chloride, followed by permeabilization with 0.2% Triton X-100 and blocking with 2% BSA in PBS (Sigma, Dorset, UK) for 1 h at room temperature. Next, cells were incubated with a primary antibody rabbit anti-lamin B1 (Abcam, Cambridge, UK); the antibody reacted both to human and rat species at 0.5 µg/mL overnight at 4 • C, followed by a secondary anti-rabbit IgG-conjugated AlexaFluor 488 (Invitrogen, Renfrew, UK) at 4.0 µg/mL for 1 h at room temperature. Finally, cells were mounted in Prolong TM gold antifade mountant (Invitrogen, Renfrew, UK). Images were acquired with a Nikon Eclipse TE2000-S fluorescence microscope (Tokyo, Japan).

Western Blot for Lamin B1
Whole-cell lysates were prepared in radioimmunoprecipitation assay (RIPA) buffer (Sigma, Dorset, UK). After discarding the nonsoluble fraction, supernatants were subjected to SDS-PAGE and immunoblot. The membranes were incubated at 4 • C overnight under agitation with a rabbit anti-lamin B1 antibody (Abcam, Cambridge, UK) diluted 1:1000 in 5% nonfat milk in Tris-buffered saline with Tween 20. Immunoblots were performed three times, quantified by ImageJ software and normalized to GAPDH.